Summary. The effect of insulin on skeletal growth was examined by (1) systemic injection, (2) local administration into the tibia growth plate and (3) in vitro by use of chondrocytes in culture.(1) Male rats, body weight 60-75 g, were hypophysectomised. One week after the operation, the animals were divided into three groups. Group A received intraperitoneally saline, group B insulin (5-30 U. kg-1. day-1) and group C human growth hormone (250 txg/kg/day) for the following 10 days. In addition, on day 10 the rats were injected with 10 txCi 35-S-sulfate intraperitoneally. Twenty-four h later in the non-fasting state plasma glucose, insulin, somatomedin activity (porcine assay), body weight, nose-rump length, width of the tibia growth plate, and the 35-S-sulfate incorporation into rib cartilage were determined. Compared to saline, growth hormone and insulin treatment significantly enhanced body weights, nose-rumb lengths, the widths of the proximal tibia growth plates and the incorporation of sulfate into rib cartilage. For the three skeletal growth parameters, growth hormone was more effective than insulin, while body weights did not differ after insulin or growth hormone treatment. Somatomedin activity (U/ml) was low in group A (0.39 + 0.04, n = 9, Mean + SEM) and group B (0.34 + 0.08, n = 8) and high in the growth hormone treated group C (0.90+0.09, n=7; p < 0.002). (2) To test the possibility that insulin might directly augment skeletal growth, insulin (80 mU) was injected into the proximal tibia growth plate of one leg and saline into the cartilage zone of the other leg. Insulin treatment significantly increased the width of the cartilage zones. Insulin: 211 + 22 ~m, saline 200 + 22 ]xm, (Mean + SD, n = 6, p < 0.05). (3) Addition of human biosynthetic insulin and growth hormone to the culture medium increased colony formation of chondrocytes in a bell-shaped fashion. A plateau in colony formation was reached with 3.1-6.25 ng/ml insulin and 25-50 ng/ml growth hormone, but with larger dosages of both hormones, the effect was diminished. The results suggest that insulin might stimulate postnatal skeletal growth by a direct effect on the target cells.